236 



PRINCIPLES OF PLANT-TERATOLOGY. 



phenomena are cited to show how the origin of the 

 normal Dicotyledonous condition may so easily have 

 come about in precisely the same way. 



The forking of the leaves in plants with opposite- 

 decussate phyllotaxis is also, on this view, a case of 

 division representing a stage towards the formation 

 of a greater number of leaves in the whorl than the 

 normal two. In the majority of cases, typified by 

 the Labiatse, the Lonicera described by Celakovsky, the 

 Buddleia and Gornus Mas (in which latter one leaf of 

 the pair had forked into three, while the other leaf was 

 separated therefrom by a brief intei node), the double 

 leaf cannot be explained by fusion, for all the other 

 leaves belonging to the nodes above and below are 

 accounted for, and there is no disturbance of the 

 phyllotaxis. A fact such as the following also supports 

 the forking theory : in a shoot of the Buddleia a whorl 

 consisting of four leaves in two opposite pairs was 

 observed ; in this case each pair was most clearly 

 derived by forking of each leaf of the pair, as the 

 phyllotaxis above and below was not disturbed, and, 

 moreover, it could not be due to the approximation of 

 two pairs of leaves, for this would involve a horizontal 

 displacement of the pairs, a phenomenon which was 

 never observed on any of the shoots. 



A case of unequivocal forking in leaves of shoots 

 with alternate phyllotaxis will now be referred to, viz., 

 in those of the weeping variety of Ulmus glabra above 

 mentioned. In this instance also the double leaves or 

 the paired leaves at each node cannot be ascribed to 

 fusion, for the phyllotaxis both above and below is 

 quite undisturbed and all other leaves are accounted 

 for. The explanation of this phenomenon, which is 

 probably the correct one, is that given by Vuillemin, 

 viz., that it represents a partial reversion to the 

 opposite-decussate type of phyllotaxis from which the 

 alternate distichous type of the normal elm-branch has 

 been derived ; the former occurs in the first vear's 

 growth of seedling elms, and may well be the original 



